D. A. Metlenkin, R. A. Platova, Y. T. Platov, M. V. Yezhova
Plekhanov Russian University of Economics, Stremyannyi per., 36, Moskva, 117997, Russian Federation
Abstract. The studies aimed to assess the information content of vegetation indices calculated from hyperspectral images of avocado fruits for their sorting according to their physiological state during storage. The work was carried out in July 2021. Four vegetation indices were used: normalized difference vegetation index (NDVI), including its modification (NDVI679/757), harvest maturity index (CFI) and plant water index (PWI). Hyperspectral images were recorded using a portable camera and spectral signatures of the avocado fruit were selected. The principal component method revealed the most informative spectral regions – 679-757 nm and 900-970 nm, which were used to calculate and modify vegetation indices. During the calculations, it was found that as the values of the PWI index increase, the NDVI and NDVI757/679 indices decrease. At the same time, the modified NDVI757/679 index has a higher inverse relationship with PWI (R2=0.99) than NDVI with PWI (R2=0.95). According to the values of two vegetation indices (PWI, NDVI679/757), using them as grayscale pseudo-colours, hyperspectral images were visualized. This made it possible to reveal a change in the physiological state of avocado fruits: darkening and an increase in the moisture content of the exocarp, which are not clearly determined by organoleptic evaluation. The proposed approach makes it possible to use hyperspectral information to determine consumer maturity and sort avocado fruits according to their physiological state during storage in commercial organizations.
Keywords: avocado; hyperspectral image; vegetation index; spectral characteristics; principal component method.
Author Details: D. A. Metlenkin, engineer (e-mail: Этот адрес электронной почты защищён от спам-ботов. У вас должен быть включен JavaScript для просмотра.); R. A. Platova, Cand. Sc. (Tech.), assoc. prof.; Y. T. Platov, D. Sc. (Tech.), prof.; M. V. Yezhova, student.
For citation: Metlenkin DA, Platova RA, Platov YT, et al. [Application of hyperspectral imaging vegetation indices for avocado fruit sorting]. Dostizheniya nauki i tekhniki APK. 2023;37(1):41-5. Russian. doi: 10.53859/02352451_2023_37_1_41.